Onto Innovation Inc. (ONTO) | The Buildout — AI Infrastructure
The Verdict
Onto Innovation builds process-control tools that inspect and measure semiconductor wafers and packages. In the AI buildout, its equipment helps manufacturers verify leading-edge logic, HBM memory, and advanced 2.5D or panel-level packages as bumps shrink and process complexity increases.
| Market Cap | — |
| Revenue (TTM) | $1.0B |
| Revenue Growth | +0.5% |
| EBITDA Margin (TTM) | 18.3% |
| Net Cash | $654M |
| Earnings Beats | 4 of 7 |
| P/E (TTM) | — |
| EV/EBITDA (TTM) | — |
What We Like
- Advanced packaging growth guidance raised from >30% to >50% in under one quarter.
- Dragonfly G5 qualified at a leading 2.5D logic customer and began shipping ahead of plan; pipeline spans over 15 applications across over 10 customers.
- Q2 2026 revenue of $343M beat the $320–330M guided range; backlog surpassed $1 billion.
- $240 million HBM volume purchase agreement covers Dragonfly 2D and 3D bump metrology through 2027, including >$60M of 3D metrology.
- Advanced nodes growth guide raised to ~25% for 2026, from a prior mid-teens range.
What We’re Watching
- The $240M HBM VPA was announced in February but was not explicitly updated on the May 5 call; watch for a shift toward a 50/50 2026/2027 mix through 2027.
- Precision optics suppliers are strained and some component lead times exceed six months; no production impact has occurred yet.
- Dragonfly G5 is still less than 10%, maybe even 5%, of 2026 advanced packaging growth; its large financial contribution depends on 2027 ramp execution.
- Power semiconductor revenue was expected to decline around 10% in 2026 and was not restated in May.
The thesis is strengthening on the evidence: growth guides were raised, Dragonfly G5 qualified ahead of plan, Q2 revenue beat the already-raised guide, and backlog passed $1 billion. The open question is whether the post-Q2 full-year guide changed and whether the 2027-specific drivers convert on time.
Earnings Beat
Q2 2026 revenue was $343 million, a record Q2 and above the $320–330 million guided range. The August 6, 2026 release highlighted strong AI-driven demand and backlog surpassing $1 billion. Full Q2 margin and segment detail were not included in the supplied source set.
| Metric | Q1 FY2026 | Q4 FY2025 | Q1 FY2025 | YoY |
|---|---|---|---|---|
| Revenue | $292M | $267M | $267M | +9.5% |
| Gross margin | 50.1% | 46.4% | 53.7% | -360bps |
| EBITDA | $64M | $40M | $76M | −15.5% |
| EPS | $0.68 | $0.21 | $1.30 | −47.9% |
| Backlog | >$1 billion | Record, ~two quarters coverage | n/a | — |
We’re, of course, thrilled to have announced our qualification adoption of Dragonfly G5 inspection system at a leading 2.5D logic customer.— Mike Plisinski, Chief Executive Officer, 2026-05-05
Management tone: Management shifted to more specific and more aggressive growth commentary than the prior call, raising advanced packaging and advanced nodes targets and confirming Dragonfly G5 qualification ahead of plan. It also gave a precise margin cadence and used direct answers in Q&A, while remaining candid about G5's small 2026 contribution and Rigaku's near-term revenue impact.
Management Guidance
On the May 5, 2026 call, management reiterated Q2 revenue of $320–330 million and guided Q2 non-GAAP gross margin of 56.0–56.5%, Q2 non-GAAP operating margin of 28.0–28.6%, and Q2 non-GAAP EPS of about $1.69 at the midpoint. Full-year guidance was set at greater than $1.3 billion, with advanced packaging growth above 50%, advanced nodes growth about 25%, and second-half revenue at least 15% above first-half revenue. Management also committed to at least 50 bps of gross margin improvement per quarter in Q3 and Q4 and an operating margin exit above 30% in Q4.
Trajectory
The trailing series is uneven: Q3 FY2025 revenue was $218M before rising to $267M in Q4 and $292M in Q1 FY2026, and Q2 2026 came in at $343M. The code-computed trajectory labels the trailing revenue series decelerating, but margin trends are expanding and the forward guide points to more than 30% FY2026 growth. The expansion is driven by advanced packaging and advanced nodes growth, Dragonfly G5 qualification and ramp, and record backlog.
The Model
The model's locked FY+1 projection is revenue of $1,350M and EBITDA of $436M, a 32.3% margin. FY+2 projects revenue of $1,780M and EBITDA of $637M, a 35.8% margin. The near-term anchor is management's FY2026 guide of greater than $1.3 billion, with advanced packaging guided up more than 50% and advanced nodes about 25%; FY+2 depends on Dragonfly G5 scale, JetStep panel-level ramp, and additional income streams from Rigaku.
| Metric | FY2025 | Next FY (E) | Following FY (E) |
|---|---|---|---|
| Revenue | $1.0B | $1.4B | $1.8B |
| YoY Growth | — | +34.3% | +31.9% |
| EBITDA | $200M | $436M | $637M |
| EBITDA Margin | 19.9% | 32.3% | 35.8% |
Projections are the median of 5 independent model runs.
On the May 5, 2026 call, management reiterated Q2 revenue of $320–330 million and guided Q2 non-GAAP gross margin of 56.0–56.5%, Q2 non-GAAP operating margin of 28.0–28.6%, and Q2 non-GAAP EPS of about $1.69 at the midpoint. Full-year guidance was set at greater than $1.3 billion, with advanced packaging growth above 50%, advanced nodes growth about 25%, and second-half revenue at least 15% above first-half revenue. Management also committed to at least 50 bps of gross margin improvement per quarter in Q3 and Q4 and an operating margin exit above 30% in Q4.
What Could Go Right — and Wrong
- Dragonfly G5 pipeline across >10 customers and >15 applications converts into named production orders.
- The HBM VPA shifts further into 2026 and is followed by additional volume purchase agreements.
- Panel-level packaging adoption accelerates beyond the roughly $200 million multi-year framework.
- Atlas G6 and critical films wins extend advanced nodes beyond memory into leading-edge logic.
- Hybrid optical/X-ray metrology moves from an N+2 R&D project toward a commercial roadmap.
- Dragonfly G5 ramp slips in 2027 or competitors win back share in the next evaluation cycle.
- The unnamed HBM customer's VPA pauses or pushes out; the absence of a May update is already a watch item.
- Precision optics lead times extend enough to delay G5 or Atlas shipments.
- AI packaging demand decelerates after the current capacity expansion.
- Rigaku fair-value accounting creates non-operating volatility without consolidated revenue, and the zero-coupon converts dilute if converted.
Looking Ahead
The next 12 months test execution more than demand: Q3/Q4 margin prints against the at least 50 bps quarterly commitment, G5 shipment cadence through Q4, and Atlas TSV shipments in 2H 2026. Into 2027, JetStep panel ramps and the HBM VPA through 2027 are the main signposts.
- Q3/Q4 2026Gross margin cadence — Tests at least 50 bps sequential improvement per quarter and Q4 exit >30% operating margin.
- 2H 2026Atlas TSV metrology shipments — Initial shipments expected, confirming the new application win.
- Q3–Q4 2026Dragonfly G5 shipment cadence — Management expects more G5 shipments each quarter through Q4.
- 2027JetStep panel-level ramp — Ramp expected at two packaging suppliers to AI device manufacturers.
- Through 2027HBM VPA execution — Tests the >$240M agreement and any shift toward 50/50 mix.
Financials
Annual Summary
| Metric | FY2024 | FY2025 | TTM | YoY |
|---|---|---|---|---|
| Revenue | $987M | $1.0B | $1.0B | +1.8% |
| Gross Margin | 52.2% | 49.8% | 48.8% | 248bps |
| EBITDA | $249M | $200M | $1.5B | -19.7% |
| EBITDA Margin | 25.3% | 19.9% | 18.3% | 535bps |
| Net Income | $202M | $137M | $106M | -32.2% |
| Free Cash Flow | $214M | $300M | $1.2B | — |
| Net Cash | — | — | — | — |
Key Ratios (Trailing)
- P/E TTM—
- EV/EBITDA TTM—
- EV/Revenue TTM—
- Price/FCF TTM—
- Gross Margin (TTM)48.8%
- EBITDA Margin (TTM)18.3%
- Net Margin (TTM)10.3%
- ROIC5.8%
- FCF Conversion126.8%
- SBC / Revenue2.7%
The Company
Onto Innovation makes semiconductor process-control systems: metrology, inspection, lithography, and data-analysis tools used by silicon wafer manufacturers, integrated circuit fabricators, and advanced packaging manufacturers. Its products matter to AI infrastructure because leading-edge logic, HBM memory, and 2.5D or panel-level packaging require tighter measurement and inspection as bumps shrink and process complexity increases.
The 10-K lists U.S. engineering, manufacturing, and service sites in Massachusetts, California, New Jersey, Minnesota, Oregon, Washington, Arizona, and Florida, plus sales and service sites in South Korea, Taiwan, China, Japan, and Singapore. Headcount was approximately 1,790 at March 31, 2026. Management has shifted capacity toward extended factories, which shipped more than 50% of Q4 2025 tools.
Business Segments
Competitive Landscape
The 10-K names KLA and Camtek Ltd. in advanced packaging inspection, Ushio and Canon in advanced packaging lithography, and PDF Solutions in software products. The source set includes neighbor disclosures from KLA and Nova that corroborate advanced-packaging expansion, but it does not map them into ONTO's formal 10-K competitor categories.
- KLANamed in ONTO's 10-K for advanced packaging inspection and mapped elsewhere in process control; neighbor disclosure says KLA's advanced packaging process-control portfolio is expected to grow to about $1B in 2026.
- Camtek Ltd.Named in ONTO's 10-K for advanced packaging inspection; neighbor disclosure records a >$260M HBM-related order/forecast for 2026–2027.
- Ushio, Inc.Named in ONTO's 10-K for advanced packaging lithography; not otherwise discussed in the source set.
- Canon, Inc.Named in ONTO's 10-K for advanced packaging lithography; not otherwise discussed in the source set.
- Named in ONTO's 10-K for software products; not otherwise discussed in the source set.
Supply Chain
ONTO sits between component suppliers and chip or package manufacturers. The source set does not record any neighbor naming ONTO by name, but adjacent equipment and OSAT disclosures corroborate the same advanced-packaging expansion.